Essentially the lensed objects show three very slightly different temporal views of the same galaxy. The time differences are pretty short in astronomical time scales - 0, +1 and +3 years (ish) - but since supernovae are extremely transient, it only appears in the T=0 image. In the crops, the supernova is marked between two horizontal bars, it's quite common in "science" images for the object of interest to be relatively unremarkable compared to the bright stuff nearby.
Thanks for the explanation about the horizontal bars! I mistakenly thought that the brighter spot on the ~1000 days later picture was the supernova and that it was mislabeled.
It's also worth considering the fact you can see a point source that's a similar brightness to the host galaxy which probably contains several orders of magnitude more stars. A Type 1A peaks at something like 5 billion times the brightness of the Sun.
Look up how black holes look like (based on simulations) for something similarly incredible. When facing the hole from a certain perspective, you can see the accretion disk from the front and the back at the same time, because the hole bends the light so much.
Essentially the lensed objects show three very slightly different temporal views of the same galaxy. The time differences are pretty short in astronomical time scales - 0, +1 and +3 years (ish) - but since supernovae are extremely transient, it only appears in the T=0 image. In the crops, the supernova is marked between two horizontal bars, it's quite common in "science" images for the object of interest to be relatively unremarkable compared to the bright stuff nearby.